4.1. Résumé
La ponction veineuse est un geste infirmier très fréquent en réanimation néonatale. Il ’e iste pas d’ tude ulti e t i ue d iva t de faço e haustive la p ise e harge antalgique de ce geste. Les objectifs étaient de décrire la fréquence de la ponction veineuse ainsi que sa prise en charge antalgique dans les services de réanimation o atale d’Île-de-France, puis de déterminer les fa teu s asso i s à l’a se e d’a alg si ue sp ifi ue à e geste ai si ue les fa teu s asso i s à u score de douleur élevé. Pour ce faire, nous avons utilisé la base de données EPIPPAIN 2 en ne sélectionnant que les nouveau-nés qui avaient subi au moins une ponction veineuse pendant la du e de l’ tude.
Sur les 589 nouveau- s i lus da s l’ tude EPIPPAIN , o t su i au oi s u e po tio vei euse et o t do t i lus da s os a al ses. L’âge gestationnel moyen (ET) était de 33,0 (4,4) semaines et la durée de participation était de 8,0 (4,5) jours. Les nouveau-nés prématurés de moins de 33 SA représentaient 51,9% de notre échantillon. La moyenne (ET ; Extrêmes) du nombre de ponctions veineuses par nouveau-né pe da t la du e de l’ tude était de 3,8 (2,8 ; 1-19) pour la totalité des nouveau-nés et de 4,1 (2,9 ; 1-17) pour les nouveau-nés de moins de 33 SA. Sur les 1 887 ponctions veineuses étudiées, 437 (23,2%) ont été réalisées pendant que le nouveau-né recevait une sédation/analgésie en perfusion continue, 1 434 (76,0%) avec un analgésique spécifique au geste. Ap s ajuste e t, l’a se e d’a alg si ue sp ifi ue au geste tait asso i e à u s o e de s v it élevé de l’ tat de sa t i itial du ouveau-né, à la p se e d’u eta d de oissa e i t a-utérin, à une ventilation trachéale ou non invasive, à la réalisation du geste la première journée d’hospitalisatio ou pendant la uit et à l’utilisatio d’u e s datio /a alg sie e o ti ue. Les scores de douleur élevés, esu s ave l’ helle DAN, taie t asso i s à l’a se e des pa e ts pe da t la po tio vei euse, à la p se e d’u e hi u gie pe da t la p iode de l’ tude et à u o e lev de tentatives pour réussir le geste.
Cette étude conclut que la ponction veineuse est un geste réalisé fréquemment chez les nouveau-nés à terme et prématurés. Plus des ¾ des ponctions veineuses sont réalisées avec une analgésie spécifique au geste. Au regard des scores de douleur, il paraît essentiel de développer des stratégies pour diminuer le nombre de tentatives et promouvoir la présence parentale lors de ce geste. Les résultats de cette étude ont été valorisés lors de congrès nationaux et internationaux ainsi que par une publication : Courtois E et al. The burden of venipuncture pain in neonatal intensive care units: EPIPPAIN 2, a prospective observational study. Int J Nurs Stud 2016 May;(57): 48–59.
The burden of venipuncture pain in neonatal intensive care
units: EPIPPAIN 2, a prospective observational study
EmilieCourtoisa,b,*,Patricia Cimermanc, Vale´rieDubuched,
Marie-FranceGoisete, ClaireOrfe`vref, AudreyLagardeg,BettySgaggeroh, Ce´lineGuioti,Me´lanieGoussotj,EtienneHurauxk,Marie-ChristineNanquettel, Ce´lineButelm,Anne-Marie Ferreiran,Sylvie Lacosteo,Sandrine Se´journe´p, Vale´rie Jollyq,Gladys Lajoier, Vale´rieMaillards,Romain Guedja,b,
He´le`neChappuya,t,Ricardo Carbajala,b,t
aEmergencyDepartment,HoˆpitalArmand-Trousseau,Paris,France
bInsermUMR1153,Obstetrical,PerinatalandPediatricEpidemiologyResearchTeam(Epope´),CenterforEpidemiologyandStatistics SorbonneParisCite´,DHURisksinPregnancy,ParisDescartesUniversity,France
cCentreNationaldeRessourcesdeluttecontrelaDouleur,HoˆpitalArmand-Trousseau,Paris,France
dNICU,HoˆpitalAndre´ Gre´goire,Montreuil,France
eNICU,Institutdepue´ricultureetdepe´rinatalogie,Paris,France
fPICU,HoˆpitalBiceˆtre,LeKremlinBiceˆtre,France
gNICU,CentreHospitalierIntercommunaldeCre´teil,Cre´teil,France
hNICU,HoˆpitalAntoineBe´cle`re,Clamart,France
iNICU,HoˆpitalRobertDebre´,Paris,France
jNICU,HoˆpitalLouiseMichel,Evry,France
kNICU,HoˆpitalCochinPortRoyal,Paris,France
lPICU,HoˆpitalArmand-Trousseau,Paris,France
mNICU,HoˆpitalLouisMourier,Colombes,France
nNICU,IntercommunaldePoissy,Poissy,France
oPICU,HoˆpitalNecker,Paris,France
pNICU,CentreHospitalierdeMeaux,Meaux,France
qNICU,CentreHospitalierDelafontaine,Saint-Denis,France
rNICU,CentreHospitalierRene´ Dubos,CergyPontoise,France
sNICU,CentreHospitalierVictorDupouy,Argenteuil,France
tUPMC,Paris,France
ARTICLE INFO
Articlehistory:
Received9November2015
Receivedinrevisedform27January2016 Accepted30January2016
ABSTRACT
Background: Newborns in intensive care units (ICUs) undergo numerous painful procedures including venipunctures. Skin-breaking procedures have been associated with adverse neurodevelopment long-term effects in very preterm neonates. The venipuncturefrequencyanditsrealbedsidepainmanagementtreatmentarenotwell knowninthissetting.
Objectives:To describe venipuncture frequency, its painintensity, and theanalgesic approach in ICU newborns; to determine the factors associated with the lack of preproceduralanalgesiaandwithahighpainscoreduringvenipuncture.
* Correspondingauthorat:ServicedesUrgencesPe´diatriques,HoˆpitalArmandTrousseau,AssistancePublique-HoˆpitauxdeParis,26AvenueduDocteur
International Journal of Nursing Studies
Whatisalreadyknownaboutthetopic?
Newborns undergo numerous painful procedures in-cludingvenipuncturesintheICU.
Skin-breaking procedures have been associated with adverse neurodevelopment long-term effects in very pretermneonates.
Analgesictreatmentsarenotalwaysusedduringpainful procedures.
Whatthispaperadds
During the first two weeks of ICU admission, very pretermneonatesundergoameanof4.1venipunctures and afourthof themundergomorethan 5 venipunc-tures.
76% of venipunctures wereperformed with preproce-duralanalgesia.
Only6/10venipuncturesweresucceededinoneattempt. Multipleattemptswereassociatedwithhigherpain. 1. Introduction
Newborns in intensive care units (ICUs) undergo numerouspainfulprocedures(Carbajaletal.,2008;Simons etal.,2003).Thisisworrisomebecauseavastliteraturehas shown that neonates are able tofeel pain (Anand and Hickey,1987;Fitzgerald,1991).Studiesusing,forexample, theflexorreflexthresholdasameasureofsensationhave evenshownthatpreterminfantsaremoresensitivetopain
2014) and motor function (Grunau et al., 2009), and impairedbraindevelopment(Anandetal.,2013;Zwicker etal.,2013);intermneonates,alteredpainresponseshave beenfound after neonatalpain (Taddio et al., 1997). It appearsthatearlypain/stressmayinfluencethe develop-ing brain and thereby neurodevelopment and stress-sensitive behaviors, particularly in the most immature neonates(RangerandGrunau,2014).Ithasthusbecome essential to prevent and treat pain in newborns. The acknowledgmentoftheimportanceofneonatalpainhas led scientific organizations to issue guidelines for the preventionof painduring invasiveprocedures by phar-macological and nonpharmacological means (AFSSAPS, 2009; American Academy of Pediatrics Committee on FetusandNewbornetal.,2006).
Venipuncturehasbeenreportedinalargemulticenter studyasthesecondmostfrequentskinbreakingprocedure inneonatesintheICU(Carbajaletal.,2008).Itisusually performedinthissettinginordertoinsertanintravenous lineortoobtainbloodforanalyses(BarkerandRutter,1995; Carbajaletal.,2008). Venipunctureisconsideredpainful (Anandetal.,2005)anditoftenrequiresmultipleattempts tobecompleted.TheEPIPPAIN1study,whichcollecteddata in2005–2006,showedthatnewbornsmighthaveasmany as 14 attemptsfor intravenous cannulainsertion before successfulcompletion(Carbajaletal.,2008).EPIPPAIN1was anepidemiologicalstudythatprospectivelycollecteddata on all painful and stressful procedures performed in neonatesadmittedtoallICUsoftheParisregion(Carbajal etal.,2008).Numerousrandomizedtrialshaveassessedthe efficacy of analgesics during venipunctures in neonates
Setting: All16neonatalandpediatricICUsintheParisregioninFrance.
Participants: AllnewbornsintheICUwithamaximumcorrectedageunder45weeksof gestationonadmissionwhohadatleastonevenipunctureduringthestudyperiod.
Methods:Dataonallvenipunctures,theirpainscoreassessedwiththeDANscaleandtheir correspondinganalgesic therapieswere prospectively collected. Theinclusion period lastedsixweeks,fromJune2,2011,toJuly12,2011.Newbornswerefollowedfromtheir admissiontothe14thdayoftheirICUstayordischarge,whicheveroccurredfirst.
Results:495newborns whounderwentvenipunctureswereincluded.Themean(SD) gestationalagewas33.0(4.4)weeksanddurationofparticipationwas8.0(4.5)days.A totalof257(51.9%)neonateswereverypreterm(<33weeks).Themean(SD;range) numberofvenipuncturesperneonateduringthestudyperiodwas3.8(2.8;1–19)forall neonatesand4.1(2.9;1–17)forneonates<33weeks.Ofthe1887venipunctures,1164 (61.7%) were performed successfully in one attempt, 437 (23.2%) with continuous analgesia,1434(76.0%)withspecificpreproceduralanalgesia.Inmultivariatemodels,lack ofpreproceduralanalgesiawasassociatedwithhigherdisease-severityscore,intrauterine growthretardation,invasiveornoninvasiveventilation,venipunctureperformedonthe firstdayofhospitalizationoratnighttime,andtheuseofcontinuoussedation/analgesia. Highpainscoresweresignificantlyassociatedwithabsenceofparentsduringprocedures, surgeryduringthestudyperiod,andhighernumberofattempts.
Conclusions: VenipunctureisveryfrequentinpretermandtermneonatesintheICUs.76% were performed with preprocedural analgesia. Strategies to reduce the number of attemptsandtopromoteparentalpresenceseemnecessary.
ß2016ElsevierLtd.Allrightsreserved.
unit Neonate Nursingcare Pain Pain management Proceduralpain Venipuncture
breastfeeding(Shahetal.,2012).Trialsofpharmacological analgesiafoundthatlidocaine–prilocainecreamwasalso effective(Biranetal.,2011;Taddioetal.,2011).Theabove nonpharmacological and pharmacological methods are usedspecificallypriortovenipuncturestoreduceneonatal pain.Continuous backgroundinfusions ofanalgesics are usedtoreducepainfrommechanicalventilation,surgeryor repeated pain (Aranda et al., 2005). However, a study showedthata continuousmorphinewasnoteffectiveto reducethepaininducedbyanotherskin-breaking proce-dure,heelstick,inpretermneonates(Carbajaletal.,2005). Notwithstandingthefactthattheseanalgesicstrategiesare knownby ICU staff, thereappears tobea gap between knowledgeandpractices.Surveyinterviewsreport substan-tialuseofanalgesicsduringvenipuncture(Lagoetal.,2013; Taddioetal.,2009)butbedsidestudiesshowthattheiruseis notroutine(BarkerandRutter,1995;Carbajaletal.,2008; Harrison et al., 2009). In the EPIPPAIN 1 study, 64% of 1333venipunctureswereperformedwithspecificanalgesia fortheprocedure(Carbajaletal.,2008).Thepublicationof theEPIPPAIN1studypresentedageneraldescriptionofthe frequencyofallpainfulandstressfulproceduresaswellas theiranalgesicmanagement;itdidnotfocusspecificallyon venipunctures.Toourknowledge,no previousstudyhas analyzedthepainmanagementtreatmentsofvenipuncture andthefactorsmodifyingit;likewise,nopreviousstudyhas analyzed the factors modifying the pain score during venipuncture.Thisinformation isessentialforimproving analgesicmanagementintheICU.
Since an important concern has been raised on the deleteriousneurodevelopmentallong-termeffectsof skin-breakingprocedures(RangerandGrunau,2014)invery pretermneonatesandsincevenipuncturesconstituteone ofthemostfrequentoftheseprocedures(Carbajaletal., 2008;Simons et al.,2003)we sought todetermine the burden of venipunctures on very sick neonates, born pretermoratterm,admittedtoICUs.Theobjectivesofthis studyweretodeterminethefrequencyofvenipunctures, the intensityof venipuncture pain in real practice, the venipuncture pain management treatments in neonates hospitalizedintheICU,thefactorsassociatedwiththelack ofspecificanalgesiauseforthisprocedure,andthefactors associatedwithahigherpainscoreduringvenipuncture. 2. Methods
2.1. Studydesign
Thepresentstudyisadetailedfocusonvenipuncturesas partofthelargeEPIPPAIN2(EpidemiologyofProcedural PainIn Neonates) study. EPIPPAIN 2 was a prospective observationalstudythatcollecteddataatthebedsideonall painfulandstressfulproceduresperformedinneonatesas well as on the pain management treatments of these
2011, to July 12, 2011. The newborns included were followed fromtheir admissionto the 14th day of their NICUorPICUstayordischarge,whicheveroccurredfirst. Onlynewlyadmittedneonateswereincluded;thosealready hospitalizedatthestartofthestudywerenotincluded.The studyincludedallpretermneonatesyoungerthan45 post-conceptional weeks and term neonates younger than 28 dayson thedayofICUadmission.Thepresentstudy focusedon neonateswho hadat leastone venipuncture duringthefirst14daysofhospitalization.Avenipuncture wasdefinedasaprocedureperformedwithaneedle,withor withoutacatheter,forobtainingintravenousaccesswith thepurposeofintravenoustherapyorbloodsampling.The only exclusion criterion was parental refusal to data collection.
2.2. Datacollection
Nursingandmedicalstaffcollecteddatainrealtime,24h adayatthebedside,onstandardizeddatacollectionforms that were introducedfor the study only. To ensure the participationofallhealthcareproviders,apresentationof thestudyandatrainingonhowtofillintheformswere conducted in each unit one monthbefore the inclusion period.Medicalandnursecoordinatorswereappointedin eachunittohelpICUstaffandtocommunicatewiththe principalinvestigators.Thedatacollectionformwasusedto recordbaselinecharacteristicsoftheneonatesincluding,but not restricted to, gestational age at birth, birth weight, intrauterinegrowthrestriction,sex,inborn,ageat admis-sion,Apgarscoreat1and5min,ClinicalRiskIndexforBabies (CRIB)score,surgeryduringthestudy period,durationof participation,andstatusatdischarge.Wealsorecordeddata on thetypeofventilationandthe useofanycontinuous analgesiaand/orsedation.Wedidnotcollectdataonthe reasonsfortheuseofcontinuousanalgesiaand/orsedation. Dataontreatmentsnotrelatedtoanalgesia/sedationwere not collected. Conditions related to each venipuncture (number of attempts, type of staff members performing the procedure, parents’ presence, hour of the day), the bedside pain assessment and the specific preprocedural analgesia administered were also recorded by the staff participatingattheprocedure.Wedidnotcollectdataonthe experienceorspecificcharacteristicsofeachmemberofthe ICUstaffwhoperformedaprocedure.Specificpreprocedural analgesiaincludednonpharmacological(e.g.sweetsolution, sucking)orpharmacologicaltreatments(e.g.Emla1cream, opioids).Acontinuousanalgesiainfusionwasnotconsidered asaspecificpreproceduraltreatment.Thepainwasassessed bythe ICUstaff whoperformedthe procedureusingthe DouleurAigue¨ Nouveau-ne´ (DAN)behavioralscale(Biran et al., 2011; Carbajal et al., 1997). This validated scale assessesproceduralneonatalpainusingfacialexpressions, limbs movements and verbal expression (crying) (Biran
BasedonEPIPPAIN1studydata(Carbajaletal.,2008), we estimated that approximately 4500 neonates are admittedtoICUsintheParisregionannually(52weeks). TodeterminesamplesizeforEPIPPAIN2,weaimedfora 95%confidencelevel5%forallthepercentagesobtained. Sample size calculations carried out with the MaCorr Research Sample Size Calculator (www.macorr.com/ ss_calculator.htm)showedthat354newbornswereneeded tomeettheobjectivessetabove.Wethusestimatedthata 6-weekenrollmentperiodwasneededtoobtain354neonates, assumingafinalinclusionrateof70%.
2.4. Dataanalysis
Data wererecordedwith EPIDATA software(version 3.1,Odense,Denmark) andanalyzedwithSPSSsoftware (version17, Chicago,IL)and Statasoftware(version13, Stata Corporation, CollegeStation, TX, USA). A research assistant verified the completeness and accuracy of all values.
Quantitative demographic data (gestational age at birth, birth weight, age at admission, Apgar score at 1and5min,CRIBscore)aswellasdataonvenipunctures (number of venipunctures, number of attempts, pain score) are presented as means (standard deviations), medians (interquartile ranges), and ranges. Qualitative variables(typeofICU,gestationalagegroup,intrauterine growth restriction, sex, inborn/outborn status, surgery, hospitalizedfor morethan 14 days,status atdischarge, analgesic treatments, parental presence, and type of operators) are presented as the numbersof individuals withpercentages.
Weperformedunivariateandmultivariateanalysesto identify the factorsassociated withthe lack of specific preproceduralanalgesiaforvenipunctures.All character-isticsofneonatesandvenipunctureswithaPvalue<0.05 in theunivariate analyses(gestationalage group, intra-uterinegrowthrestriction,inborn/outbornstatus,surgery, type ofventilation,CRIBscore,ageatadmission, dayof procedure,time ofday,anduseofcontinuoussedation/ analgesia), and theneonate’s sex were included in the multivariatemodel.
Toanalyzetheintensityofvenipuncturepainaswellas to determine the factors associated witha higher pain score,weperformedunivariateandmultivariateanalyses. Ahighpainscorewasdefinedasascoregreaterthan3on theDANscale.Inthevalidationstudy,painfulprocedures yieldedscoresfrom1to10,with95%ofscoresgreaterthan 3(Carbajaletal.,1997).AllvariableswithaPvalue<0.05 intheunivariateanalyses(Apgarscore,typeofventilation, parental presence,surgery, number of attempts), gesta-tional age, sex, number of preceding skin-breaking procedures, and the use of specific and continuous sedation/analgesiaduringvenipuncturewereincludedin themodel.
Since venipuncturesare not independent events and
95%confidenceintervals(95%CI).BecausetheGEEmodel is based on the quasi-likelihood theory instead of the maximumlikelihoodtheory(Wedderburn,1974),weused the quasi-likelihood under the independence model criterion (QIC)tocomparedifferent GEEmodelsand to choosetheonewiththesmallestQICvalue(Pan,2001).
Statistical significance was set at P<0.05 for all analyses.
2.5. Ethicalconsiderations
The local committee for the protection of human subjects determined that further approvals or parental consentfordatacollectionswerenotrequiredaccordingto Frenchlawbecausethiswasanobservationalstudywith nochangesinthestandardofcare.Parentswereinformed aboutthestudyandcouldrefusedatacollectionfortheir infant(s).TheFrenchDataProtectionAuthorityapproved the computerized data collection. The study was registeredbeforedatacollectionontheClinicalTrials.gov (NCT01346813).
3. Results
3.1. Populationcharacteristics
Fromthe625newbornsfulfillingtheinclusioncriteria, 589 were included in the EPIPPAIN 2 study (Fig. S1, Supplementary material). From these,495 (84.0%)from the16centersunderwentatleastonevenipunctureand wereincludedinthisstudy.
In2011,theParisRegionregistered11852851 inhabi-tants and 182284 births per year. The corresponding figuresforallFrance(metropolitan)were64933400 inha-bitantsand821589>birthsperyear(INSEE,2011).
Themean (SD)gestationalagewas 33.0(4.4)weeks gestation.From the495infants, 389(78.6%) wereborn preterm(<37wks)and 257(51.9%)were verypreterm (<33wks).Themean(SD)durationofparticipationwas8.0 (4.5)calendardaysandtheobservationperiodrepresented 4372 patient-days. Table 1 summarizes the principal characteristicsofthestudypopulation.
3.2. Characteristicsofthevenipunctures
We identified a total of 1887 venipunctures: 1152 (61.0%)foronlybloodsamplingand735(39.0%)forthe placementofanintravenousline(withorwithoutblood sampling). During the study period, the mean (SD) numberofvenipuncturespernewbornwas3.8(2.8),the median(IQR)was3.0 (2.0–5.0) witharangefrom1 to 19perinfant.Inthegroupofinfantsborn<33weeks,the mean(SD),themedian(IQR),andtherangenumbersof venipunctures were 4.1 (2.9), 4.0 (2.0–5.0), and 1–17. The neonate who underwent 19 venipunctures did so
Fig.S2(Supplementarymaterial)showsthefrequency distributionofvenipuncturesbygestationalageatbirth.Of the 1887venipunctures,437(23.2%) occurredwhilethe neonatewasreceivingcontinuousanalgesiainfusionsand 1434(76.0%)occurredwithspecificpreprocedural analge-sia.Overall, 1639(86.9%)venipunctureswere performed withsomeanalgesictreatment,continuousinfusionsand/or specific preprocedural analgesia. A total of 436(23.1%) venipunctureswereperformedwhentheneonateswereon continuousopioidinfusions.Fromthe1887venipunctures, 1164(61.7%),343(18.2%),188(10.0%)and192(10.2%)were successfullyperformedinoneattempt,2,3and4ormore attempts,respectively.Correspondingfiguresfor1056 veni-punctures performed in infants <33 weeks were 681 (64.5%),169(16.0%),111(10.5%)and95 (9.0%).Fromall
of 1337 (70.9%) venipunctures were performed with nonpharmacologicalanalgesiaincluding1106(58.6%)that were performed with sweet solutions, of which 753 (39.9%)were withglucose,294(15.6%)sucrose, and 59